• Typically meant to be shared for a larger audience (usually through html, RMarkdown, Shiny)
• Enable customized data exploration experience through linking/interactivity
• Review findings in and across multiple dimensions
• Visualize data trending through animation

## [1] "AAPL" "MSFT"

• Define the message you want to communicate (simple and specific)
• Structure your visualization
  • Macro view: the entire set of data
  • Micro view: detailed information/ explanations of a single data point
  • Customized view: customizations enabled in your plot
• Add Interactivity
  • Choose gestures
  • Define trigger behaviors
  • Animated elements (reveal trending)
• Stylize

Plotly is a JavaScript-based library that enables interactive plotting, and it also includes a ggplotly() function to turn graphs created with ggplot2 interactive.

d <- diamonds[sample(nrow(diamonds), 1000), ] 
p <- ggplot(data = d, aes(x = carat, y = price)) +
  geom_point(aes(text = paste("Clarity:", clarity))) + facet_wrap(~ cut)
ggplotly(p,height=350)

• https://plot.ly/r/reference/ https://plot.ly/python/reference/

df <- read.csv("https://raw.githubusercontent.com/plotly/datasets/master/2011_us_ag_exports.csv")
df$hover <- with(df, paste(state, '<br>', "Beef", beef, "Dairy", dairy, "<br>",
                           "Fruits", total.fruits, "Veggies"))
l <- list(color = toRGB("white"), width = 2)
g <- list(scope = 'usa', projection = list(type = 'aitoff'), showlakes = TRUE)
plot_ly(df,  type = 'choropleth', z = ~total.exports, text = ~hover, 
    locations = ~code, locationmode = ~'USA-states', # map state code to state locations
    marker = ~list(line = l), 
    color = ~total.exports, colorbar = ~list(title = "Millions USD"),
    height=300) %>%
    layout(title = '2011 US Agriculture Exports by State', geo = g)

• Need to install rCharts from github using the devtools package
• Supports multiple javascript charting libraries
• Share through chart$publish('mychart', host = 'rpubs') or Shiny or html NVD3

require(rCharts)
data(economics, package = 'ggplot2')
ecm <- reshape2::melt(economics[,c('date', 'uempmed', 'psavert')], id = 'date')
p <- nPlot(value ~ date, group = 'variable', data = ecm, type = 'lineWithFocusChart')
p$set(height=200)
p$print('mychart', include_assets = TRUE, cdn = TRUE)

Highcharts

h1 <- Highcharts$new()
h1$chart(type = "spline")
h1$series(data = c(1, 3, 2, 4, 5, 4, 6, 2, 3, 5, 4, 7), dashStyle = "longdash", name='2011')
h1$series(data = c(NA, 4, 1, 3, 4, 2, 9, 1, 2, 3, 4, 3), dashStyle = "shortdot", name='2012')
h1$series(data = c(4, 6, 7, 8, 6, 2, 9, 4, 3, 6, 7, NA), dashStyle = "longdot", name='2013')
h1$xAxis(categories = 1:12); h1$legend(symbolWidth = 80)
h1$set(height = 300)
h1$print('mychart2', include_assets = TRUE, cdn = TRUE)

A grammar of graphics provides a structure to combine graphical elements into figures that display data in a meaningful way

Building blocks of a graph include:

• data
• coordinate system: cartesian, polar
• geometric object: marks you used to represent data (points, line, bins)
• aesthetic mapping: visual property of the objects in the plot, anything you can see (location, shape, size)
• statistical transformations: mean, median

• data, coordinate system, geometric object, aesthetic mapping, statistical transformation

ggvis combines: 1) a grammar of graphics from ggplot2 2) reactive programming from shiny 3) data transformation pipelines from dplyr

library(ggvis)

mtcars %>%
  ggvis(x = ~wt, y = ~mpg, size.hover := 200,
        stroke := NA, stroke.hover := "red", strokeWidth := 3) %>%
  layer_points() %>%
  set_options(height=300)

All of them

• based on JavaScript
• compatible with RMarkdown
• in an early development phase (limited functionality and control over graphical elements)

• plotly – convert ggplot2 figures to interactive plots easily
• rCharts – R interface to multiple javascript charting libraries; syntax used varies between different types of charts
• ggvis – interactive plots from the makers of ggplot2, use %>%, does not support map

• googleVis – use Google Chart Tools from R, well-documented, relatively mature
• d3Network package - cool interactive network visualizations